The efficacy of fenfluramine in infantile autism has received recently a great deal of publicity. It appears to be a promising treatment in this population. However, well designed research and careful monitoring in an inpatient setting is required in order to determine the role of this drug. We wish to assess critically the short- and long-term efficacy and safety of fenfluramine and its effects on learning in autistic children, and to compare them to those of haloperidol and placebo. Haloperidol is chosen as a control drug, because its effects on behavioral symptoms and on learning and its safety were systematically investigated for some time in our unit: we have demonstrated that it is both statistically and clinically superior to placebo in autistic children when administered in well designed studies under double-blind conditions. The design we have chosen is such, that each child will serve as his own control. The children (N=42; ages 2 to 7 years) will be carefully diagnosed using DSM-III criteria for Infantile Autism, Full Syndrome Present; those with known causes for autism (Axis III) will not be included in the experiment, thus increasing the homogeneity of the sample. The short-term study will be over a period of 14 weeks and the long-term study over 3 years. In both short and long-term studies, tasks, measures, strategies and scales will be used as in our previous research with haloperidol. We have the expertise and experience to conduct this type of research, the population and the required facilities which include an automated laboratory. We also have shown in the past that we are able to achieve 100% drug compliance and have the ability to complete a study on time, as scheduled. Comparing 2 drugs of different actions, such as fenfluramine and haloperidol, may aid us in delineating responders to each and, thus, to shed some light on this heterogenous syndrome. This project also represents a major contribution to research in drug-induced dyskinesias. It is the only prospective project to study these abnormal involuntary movements in a population which has a high baseline rate of abnormal movements unrelated to long-term drug administration.